Misc True to Life AI 2013 Season 1.4

[snthennumbers]

but I'll run these tests tomorrow for you
looking forward to the results

when I used the primes instead, after 7-8 laps (in which I was NOT involved in any battle and was driving mainly by myself) I started to struggle, until lap 26 where I got overtaken by bottas (impossible to keep him back) and I had to repeatedly block Maldonado. I finished 11th.
so this time options were doing pretty fine, and primes were terrible.
this is also probably somewhat due to the AI tending to run mix 3 more frequently in the closing stages of most races.

ps. I also have to admit that I never raced a 100% race with 2012 before... so maybe I don't have clear how codemasters thought the tyres were supposed to react... but still, no difference between 50% database and vanilla database, for the options at least...
I don't do 100% races either lol. If I remember right I think the medium compound tyre was the only tyre i didn't adjust simply because I couldn't find a balance of tyre longevity for all compounds without leaving the mediums alone. Super Softs, Softs, and Hards I made some adjustments to in order to match the 2013 season spec tyre.

Remember that as your career progresses and the cars develop, you'll have some races where you struggle, and some races where you do well. I've also just personally found that the Sauber is all around a quick car in general. I did shockingly well with it in career season 1. Whereas some other people drive for Williams or Caterham in season 1 and typically don't find themselves higher than 8th or worse.

 

[snthennumbers]

Hi snthennumbers, thanks for the quick reply.

Yes, I read the discussion on the grip levels I use option tyres in Q1 and do my setups trying to increase grip as much as possible


I'm using the 2013 version. I know all levels have been increased but the real problem is that, for my skill, Intermediate level it's a bit easy and Professional it's impossible (at least with the nerfed Williams)



I see your point and partly agree with it. I know it might be hard to get the AI to do real-life-like lap times but my point was that in the Professional level they're considerably lower and seem impossible to reach (maybe this is related with the grip discussion as well.

Thanks a lot for your work once again. If you have any suggestion please let me know.

This is on the bottom of the overview page, but for the record to individualize your difficult levels:
Open the “track_difficulties” table in the database. This is on a per-track basis and purely on a personal level. Go to the “track_difficulties” table. Expert = Legend, Hard = Professional, Medium…etc etc. The “track” column corresponds to the “track_id” column in the “track_model” table (way towards the right). 1.0 is as fast they can go. You can go out to something like 6 decimal places for track difficulty.

So for you, the easiest thing to do would be to keep an excel sheet with your personal levels for the track_difficulties table. Then as I push out new versions of the mod, you can simply copy and paste in your personal values into the database and you're off and running again with your personal settings

 

[Warnek]

This is on the bottom of the overview page, but for the record to individualize your difficult levels:
Open the “track_difficulties” table in the database. This is on a per-track basis and purely on a personal level. Go to the “track_difficulties” table. Expert = Legend, Hard = Professional, Medium…etc etc. The “track” column corresponds to the “track_id” column in the “track_model” table (way towards the right). 1.0 is as fast they can go. You can go out to something like 6 decimal places for track difficulty.

So for you, the easiest thing to do would be to keep an excel sheet with your personal levels for the track_difficulties table. Then as I push out new versions of the mod, you can simply copy and paste in your personal values into the database and you're off and running again with your personal settings

Thank you! This is really appreciated

 

[AceSCMF]

Ok here's what I found out!!

I raced Melbourne, quick race, 50% race with 50% database first, and then 100% with 100% database (I'll run vanilla tomorrow maybe).

I screenshotted around the same laps, so here we go:

on the left you can see the 50% race (confirmed by the number of total laps) and on the right the 100%.
I didn't post any laptime because I was worried by taking screenshot with the keyboard on my lap driving with my speedwheel, so it's been messy enough and I couldn't make any consistent racing. But I still have the feedback of how I felt the track.

Number one, the 100% race was waaaay more slow and bulky, and tyres started degrading (as you can see) a bit earlier than 50%.
This tells me that there is no fuel scale at all (I'm talking about Codies ofc, not you )

Number two, there was absolutely no difference in the feeling on 50% and 100%, and the next screenshots will support me. This means that the tyres scaling only works suggesting you and the AI the strategy, but there is no difference at all in the wear.

This was lap 5, where I was suggested by the system to do the first stop in 50%.

this is instead lap 9, where I was suggested to do the pit in the 100% race. as you can see the tyres are way more degraded than lap 5 on 50%.

This is lap 11. Everything's fine so far. car is reacting great in both races.

This is lap 15 (lol I missed a shot actually ) car was just now starting to feel a little unconfortable. probably the times were already dropping a couple of laps before, telling me that the target lap for the supersofts was around 12.

This is lap 18. In both races the car has started to act crazy, and in 100% race, since the car was heavier, just a second after the shot I punctured my back left tyre just on the straight!
Flashbacked and went back on track.

Lap 20 was like running on a track that was starting to get wet: no traction on exit of the corners, lot of spinning, problems everywhere. On 50% I punctured at the end of the lap, on 100% I kept on sliding on half of the corners and ended on the walls several times.

So the conclusion is (at least for my installation):

- There is no tyre wear scaling, except for the suggested strategy
- There is no fuel scaling

My question now is:

is it just me that screwed up the installation, or is it happening to everybody? Can somebody run another test like this on his own installation?

 

[snthennumbers]

So the conclusion is (at least for my installation):

- There is no tyre wear scaling, except for the suggested strategy
- There is no fuel scaling

My question now is:

is it just me that screwed up the installation, or is it happening to everybody? Can somebody run another test like this on his own installation?

Well hot damn, nice work buddy. Very strange performance from the game but good find
Obviously in the database I'm making the tyres table (aka the tyre longevity) half of the default values from the 100% database. However I must need to make the tyre wear rates of each team twice as much for the 50% database (the f1teams table). As well as check the fuel levels for the 50% database.

I'm absolutely swamped with real life right now (moving apartments, and a few huge projects at work).
Do you think you could possibly test the following:

Take the 50% database > f1teams table
find the "front_tyre_wear" and "rear_tyre_wear" columns.
Double the values for each team.
Go do a 50% race and see what happens with both you and the AI (see if they're pitting oddly, or at all, etc)
Also if you like (but do the above first, then the following one at a time to allow you to see the results of each change), check the track_fuel_distance table and double the "fuel_usage_per_lap" values for each track.
If that creates odd behaviour, try going to the track_model table and cutting in half the values for the "fuel_tank_size" column. remember to do one change at a time, otherwise it's going to be really tough to determine what's doing what.

We now know that the tyre longevity adjustments (from the tyres table) are working for 100% races (though you are able to make your tyres last a long time, I think you're fairly easy on them)...but that's neither here nor there since we're focusing on scaled tyre wear for 50% races right now.

Let me know (assuming you're willing to test the above of course).
Thanks mate, great work

 

[AceSCMF]

absolutely!!

 

[w1n1x]

snthennumbers
Found one problem.I installed your latest version, but correct me if i'm wrong but the names are messed up.For example kovalainen is in williams , and timo glock is still in marussia.What i did with my game (installed HD cars , some of them are 2xHD, driver suits, helmets to correct drivers, for example:timo glock driver helmet replaced with Bianchi's; steering wheels , suits,electronic flags, all tracks update hd, and sweet fx.So what's the problem?

 

[AceSCMF]

Well hot damn, nice work buddy. Very strange performance from the game but good find
Obviously in the database I'm making the tyres table (aka the tyre longevity) half of the default values from the 100% database. However I must need to make the tyre wear rates of each team twice as much for the 50% database (the f1teams table). As well as check the fuel levels for the 50% database.

I'm absolutely swamped with real life right now (moving apartments, and a few huge projects at work).
Do you think you could possibly test the following:

Take the 50% database > f1teams table
find the "front_tyre_wear" and "rear_tyre_wear" columns.
Double the values for each team.
Go do a 50% race and see what happens with both you and the AI (see if they're pitting oddly, or at all, etc)
Also if you like (but do the above first, then the following one at a time to allow you to see the results of each change), check the track_fuel_distance table and double the "fuel_usage_per_lap" values for each track.
If that creates odd behaviour, try going to the track_model table and cutting in half the values for the "fuel_tank_size" column. remember to do one change at a time, otherwise it's going to be really tough to determine what's doing what.

We now know that the tyre longevity adjustments (from the tyres table) are working for 100% races (though you are able to make your tyres last a long time, I think you're fairly easy on them)...but that's neither here nor there since we're focusing on scaled tyre wear for 50% races right now.

Let me know (assuming you're willing to test the above of course).
Thanks mate, great work

Ok, I've run the first test.

Installed the 50% database, doubled the tyre consumption, got into the game.
Suggested pit strategy: NONE. I couldn't even add a new stop! I could only change the starting compound.

Started the race, after 10 laps I had the same wear as before.

 

[AceSCMF]

I've run the second test, with the fuel usage per lap, and it's just messing up the fuel indicator. every car starts with a -9 laps of fuel (I guess it's more but -9 is the max it can show), so every car (including me) went on mix 1. Almost at 3/4 of the race, I finally got optimal fuel, so I put on mix2. After a lap I got +1 laps so I put on mix3. But even with mix 3 I kept on "growing" the fuel... I finished the race with +5 laps and mix3.

So I guess even fuel consumption is hard coded.

I'll test tank size later.

 

[ParkYongLee]

So, it was never just my imagination, thinking that tyre wear only changes the pit stop strategy but doesn't really affect ingame tyre wear. No wonder I do 1 stop less than the AI as I change when I "feel" it's time to but the AI changes when the DB tells them to do so.

 

[Elliotlols]

To get this to work on career mode do install the 50% database and choose 50% race option?

Also, do AIs still retire from engine problems etc. during races on this mod?

 

[Tom Vaughn]

Hey guys.
Coming from version 3.1/2012version can i just copy /paste the 3.2 update or would i have to restart my career?

 

[snthennumbers]

To get this to work on career mode do install the 50% database and choose 50% race option?

Also, do AIs still retire from engine problems etc. during races on this mod?

Correct - 50% database means you can do 3 lap, 5 lap, 25%, and 50% distance races only (in career or quickrace). 100% database means you can do any distance you like. You cannot do 100% races with the 50% database otherwise the AI will never pit, causing big problems.

AI will retire from races with engine failures, etc. Usually 3 to 5 cars per race will be a DNF.

 

[snthennumbers]

Hey guys.
Coming from version 3.1/2012version can i just copy /paste the 3.2 update or would i have to restart my career?

You can copy & paste the v3.2 database without having to restart your career. For both 2012 and 2013 versions of my mod.

 

[snthennumbers]

snthennumbers
Found one problem.I installed your latest version, but correct me if i'm wrong but the names are messed up.For example kovalainen is in williams , and timo glock is still in marussia.What i did with my game (installed HD cars , some of them are 2xHD, driver suits, helmets to correct drivers, for example:timo glock driver helmet replaced with Bianchi's; steering wheels , suits,electronic flags, all tracks update hd, and sweet fx.So what's the problem?

Not a problem with my mod bud, that's an installation issue with v1tek's 2013 season mod. My mod doesn't handle any graphics files (excluding the Lotus livery change for the 2013 Lotus cars).
Check v1tek's installation guide here:
http://www.racedepartment.com/forum/threads/f1-2013-season-mod.68460/page-7#post-1426936

 

[snthennumbers]

I've run the second test, with the fuel usage per lap, and it's just messing up the fuel indicator. every car starts with a -9 laps of fuel (I guess it's more but -9 is the max it can show), so every car (including me) went on mix 1. Almost at 3/4 of the race, I finally got optimal fuel, so I put on mix2. After a lap I got +1 laps so I put on mix3. But even with mix 3 I kept on "growing" the fuel... I finished the race with +5 laps and mix3.

So I guess even fuel consumption is hard coded.

I'll test tank size later.

Boy this is interesting isnt it...so far it sounds like tyre wear for the Player is hard coded in the exe since apparently no matter what we change we're still seeing the same wear levels? However when I make slight adjustments to the front/rear_tyre_wear values in the f1teams table (like I did for Force India) they (the AI) do in fact do longer stints on their tyres.
Maybe instead of being brash and doubling the front/rear_tyre_wear values in the f1teams table, we might just need to finesse them up a bit from their default values. So for example instead of 0.97, we make it 1.17 or so...

The concept of basically cutting in half the tyre longevity values in the tyres table does (to me) make sense to do, but it sounds like the tyres table values and front/rear_tyre_wear values in the f1teams table need to be balanced together in order to get actual scaled tyre wear...

Very strange stuff, but excellent work mate. Much appreciated!

 

[robbe8]

Yeah I did try changing to "PlayerOnly=False" but didn't notice a difference. I'll try making some new coords and see where that gets me. Stuck at work right now so gotta wait till I get home :\

Gigisan1987 I actually haven't tried champions mode, but I don't see why it wouldn't work. Are you getting an error? Or does it crash to desktop? Anyone else having this issue?

I'm having the same problem, when I click on championship mode the game crashes to destkop.

EDIT: it crashes when i select championship mode and then hard difficulty, if I select medium difficulty it doesn't crash.

 

[snthennumbers]

I'm having the same problem, when I click on championship mode the game crashes to destkop.

EDIT: it crashes when i select championship mode and then hard difficulty, if I select medium difficulty it doesn't crash.

Must be an issue in the database, but its not technically my database, it's the database from v1tek's 2013 Season Mod with my modifications to driver and car performance. So that'd be a question for v1tek on his thread

 

[robbe8]

ok thanks!

 

[Rafael Silva Pereira]

Is there another link to download the True to Life V3.2? I can't download from direct link